mridulm commented on code in PR #36512:
URL: https://github.com/apache/spark/pull/36512#discussion_r873204359
##########
core/src/main/scala/org/apache/spark/storage/BlockManager.scala:
##########
@@ -933,10 +933,29 @@ private[spark] class BlockManager(
})
Some(new BlockResult(ci, DataReadMethod.Memory, info.size))
} else if (level.useDisk && diskStore.contains(blockId)) {
- try {
- val diskData = diskStore.getBytes(blockId)
- val iterToReturn: Iterator[Any] = {
- if (level.deserialized) {
+ var retryCount = 0
+ val retryLimit = 3
Review Comment:
My main concern is, usually with bad disks/etc, the reads can take an
inordinate amount of delay (due to various layers down retrying, trying to
recover) - so a read which should typically take a few ms can go into minutes
or higher: hence why I want to understand how to estimate/configure this.
One option is to make it a private config and make it user configurable -
with 3 (or 2 ?) as the default. Thoughts ?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
